rtdetr-v2-r50-finetune-26

This model is a fine-tuned version of PekingU/rtdetr_v2_r50vd on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 6.0635
Map: 0.5527
Map 50: 0.8316
Map 75: 0.6801
Map Small: 0.5136
Map Medium: 0.6695
Map Large: -1.0
Mar 1: 0.3398
Mar 10: 0.6634
Mar 100: 0.7117
Mar Small: 0.682
Mar Medium: 0.7874
Mar Large: -1.0
Map Artemia: 0.5527
Mar 100 Artemia: 0.7117

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 5e-05
train_batch_size: 8
eval_batch_size: 8
seed: 42
optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
lr_scheduler_warmup_steps: 300
num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Map	Map 50	Map 75	Map Small	Map Medium	Map Large	Mar 1	Mar 10	Mar 100	Mar Small	Mar Medium	Mar Large	Map Artemia	Mar 100 Artemia
No log	1.0	250	12.2175	0.2509	0.4704	0.262	0.1874	0.4289	-1.0	0.2888	0.5826	0.6548	0.586	0.7489	-1.0	0.2509	0.6548
217.6859	2.0	500	8.9410	0.434	0.8083	0.4068	0.3634	0.5374	-1.0	0.3368	0.5841	0.6243	0.5823	0.6832	-1.0	0.434	0.6243
217.6859	3.0	750	7.7992	0.4882	0.8365	0.4996	0.4175	0.5868	-1.0	0.3748	0.6143	0.6539	0.6016	0.7263	-1.0	0.4882	0.6539
14.0835	4.0	1000	7.6803	0.4744	0.8436	0.4646	0.3994	0.5842	-1.0	0.3782	0.615	0.6598	0.6145	0.7234	-1.0	0.4744	0.6598
14.0835	5.0	1250	8.5616	0.4666	0.8369	0.4766	0.3935	0.5661	-1.0	0.3592	0.5991	0.6153	0.564	0.6861	-1.0	0.4666	0.6153
12.553	6.0	1500	7.8614	0.4656	0.8339	0.4599	0.3808	0.5949	-1.0	0.3654	0.5769	0.596	0.5296	0.6876	-1.0	0.4656	0.596
12.553	7.0	1750	7.9009	0.4524	0.8396	0.4319	0.3688	0.5708	-1.0	0.3555	0.5773	0.6003	0.5253	0.7036	-1.0	0.4524	0.6003
11.5619	8.0	2000	8.5578	0.4559	0.8399	0.453	0.3735	0.5685	-1.0	0.3614	0.586	0.5969	0.5376	0.6788	-1.0	0.4559	0.5969
11.5619	9.0	2250	8.0713	0.4558	0.8497	0.4446	0.3712	0.582	-1.0	0.3676	0.5807	0.6078	0.5366	0.7058	-1.0	0.4558	0.6078
10.6127	10.0	2500	7.6639	0.4688	0.8498	0.4553	0.3903	0.5858	-1.0	0.3657	0.5872	0.5991	0.5317	0.6927	-1.0	0.4688	0.5991
10.6127	11.0	2750	8.5195	0.4588	0.8629	0.4532	0.3787	0.5763	-1.0	0.3617	0.5779	0.591	0.5172	0.6934	-1.0	0.4588	0.591
9.9482	12.0	3000	8.3140	0.4533	0.8498	0.427	0.3669	0.5885	-1.0	0.3639	0.571	0.5769	0.4914	0.6942	-1.0	0.4533	0.5769
9.9482	13.0	3250	8.5309	0.4632	0.8654	0.4495	0.3856	0.5758	-1.0	0.3589	0.5822	0.5916	0.5274	0.6803	-1.0	0.4632	0.5916
9.2524	14.0	3500	8.3882	0.4532	0.8479	0.4241	0.3795	0.5716	-1.0	0.357	0.5785	0.5838	0.514	0.6788	-1.0	0.4532	0.5838
9.2524	15.0	3750	8.2529	0.4488	0.8319	0.4273	0.3757	0.5621	-1.0	0.3561	0.5676	0.5713	0.5005	0.6686	-1.0	0.4488	0.5713
8.7231	16.0	4000	8.4274	0.448	0.8376	0.4272	0.3654	0.5741	-1.0	0.3604	0.5654	0.5688	0.4871	0.6818	-1.0	0.448	0.5688
8.7231	17.0	4250	9.0475	0.4557	0.8523	0.4266	0.3696	0.5746	-1.0	0.3626	0.5679	0.5704	0.4871	0.6854	-1.0	0.4557	0.5704
7.9307	18.0	4500	8.7691	0.4581	0.8595	0.4493	0.3835	0.566	-1.0	0.3583	0.5732	0.5735	0.5022	0.6723	-1.0	0.4581	0.5735
7.9307	19.0	4750	8.4439	0.4677	0.8477	0.4717	0.3873	0.5822	-1.0	0.3639	0.571	0.5717	0.4935	0.6796	-1.0	0.4677	0.5717
7.3944	20.0	5000	9.4234	0.4603	0.8542	0.4279	0.3866	0.5688	-1.0	0.3673	0.5726	0.5748	0.5032	0.6737	-1.0	0.4603	0.5748
7.3944	21.0	5250	9.3303	0.4624	0.8589	0.44	0.3824	0.5809	-1.0	0.3598	0.572	0.5769	0.4973	0.6869	-1.0	0.4624	0.5769
6.8855	22.0	5500	9.4245	0.4568	0.8433	0.4427	0.383	0.5658	-1.0	0.3595	0.5648	0.566	0.493	0.6672	-1.0	0.4568	0.566
6.8855	23.0	5750	9.2134	0.4602	0.8477	0.4368	0.3865	0.5649	-1.0	0.3601	0.5639	0.5642	0.4876	0.6701	-1.0	0.4602	0.5642
6.3313	24.0	6000	9.8657	0.4617	0.8503	0.4277	0.3922	0.5623	-1.0	0.3614	0.5723	0.5729	0.5005	0.673	-1.0	0.4617	0.5729
6.3313	25.0	6250	10.1977	0.4577	0.8498	0.4226	0.3814	0.5701	-1.0	0.3579	0.5695	0.5695	0.4952	0.6723	-1.0	0.4577	0.5695
5.906	26.0	6500	9.9314	0.4596	0.8443	0.4351	0.377	0.5711	-1.0	0.3611	0.5645	0.5645	0.4914	0.6657	-1.0	0.4596	0.5645
5.906	27.0	6750	10.0074	0.4595	0.84	0.4376	0.3825	0.5654	-1.0	0.3607	0.5688	0.5692	0.4978	0.6679	-1.0	0.4595	0.5692
5.4679	28.0	7000	10.6859	0.4631	0.8458	0.4388	0.3851	0.5687	-1.0	0.3667	0.5707	0.5707	0.5	0.6686	-1.0	0.4631	0.5707
5.4679	29.0	7250	10.2397	0.4606	0.8407	0.4439	0.3824	0.5709	-1.0	0.3642	0.5698	0.5698	0.4968	0.6708	-1.0	0.4606	0.5698
5.1029	30.0	7500	10.4180	0.4565	0.8387	0.4203	0.3805	0.5653	-1.0	0.3617	0.5698	0.5698	0.4978	0.6693	-1.0	0.4565	0.5698

Framework versions

Transformers 4.57.1
Pytorch 2.8.0+cu128
Datasets 4.2.0
Tokenizers 0.22.2

dariacuna/rtdetr-v2-r50-finetune-26

rtdetr-v2-r50-finetune-26

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

为此模型做种